Our previously reported experience of treating Pompe mice (an acid alpha glucosidase knockout strain) and the experience in trials in humans with enzyme replacement therapy (ERT), demonstrated that the drug is highly effective in clearing glycogen stored in cardiac muscle cells and poorly effective for skeletal muscle cells. This failure of skeletal muscle response to recombinant enzyme in Pompe Syndrome has become a major focus of our attention and has led us into active new areas of cell biology that in themselves add great interest to our activities. In Pompe fast fibers there are large areas of autophagic build-up, which may occupy close to half the diameter of a fiber, are usually centrally located, and disrupt the microtubular network and the contractile apparatus itself. The disturbed clearance of autophagic debris in fast fibers is accompanied by clear evidence in fast fibers of up-regulation of the autophagic pathway: increased levels of both LC3-I and Lc3-II, and increased levels of Beclin-1 and Atg7. To study this further, we made autophagy-deficient Pompe mice. For studies of ERT, two and a half month-old GAA -/-, MLCcre: Atg7F/F: GAA-/-, and HSAcre: Atg5F/F: GAA-/- mice received 3 intravenous injections of recombinant human acid alpha-glucosidase at a dose of 100 mg/kg every other week and the mice were sacrificed 7 days after the last injection. As expected, autophagy was completely suppressed in fast (gastrocnemius) but not slow (soleus) muscles in the MLCcre: Atg7F/F: GAA-/- mice as shown by the absence of LC3II, a highly specific marker for autophagic vesicles (autophagosomes). Expansion of lysosomes, a hallmark of Pompe disease, persists in muscle fibers from MLCcre: Atg7F/F: GAA-/- mice, but autophagic accumulation is absent. A prominent feature of MLCcre: Atg7F/F: GAA-/- mice is that they show an age-dependent accumulation of ubiquitinated proteins in their skeletal muscles, suggesting a functional impairment of the lysosomes. The characteristics described above are similar to those seen in the previously described HSAcre: Atg5F/F: GAA-/- mice. The level of glycogen in MLCcre: Atg7F/F: GAA-/- was lower than in the GAA-/- by 57%, suggesting that autophagy may play a role in the delivery of lysosomal glycogen in this autophagy-deficient strain. (This conclusion, based upon observations in a much larger number of both Atg5 and Atg7 autophagy knockout animals, revises our earlier conclusion about the limited role of macroautophagy in the delivery of glycogen to lysosomes.) Considering the relatively low glycogen load in MLCcre: Atg7F/F: GAA-/- mice and the lack of additional clinical manifestations, these mice were good candidates for ERT. In these mice ERT resulted in a dramatic reduction in the glycogen level approaching wild type levels. Glycogen clearance was also demonstrated by PAS staining of muscle biopsies and by immunostaining of isolated single fibers for autophagosomal and lysosomal markers. In contrast, GAA-/- mice with genetically intact autophagy cleared glycogen poorly. The same ERT regimen in the HSAcre: Atg5F/F: GAA-/- strain with higher initial glycogen levels similarly led to complete removal of the accumulated glycogen. The results suggest that the removal of autophagic buildup is a major factor that causes this therapy to be so effective. Furthermore, ERT-treatment in autophagy-deficient GAA-/- strains resulted in a significant decrease in the amount of ubiquitinated-proteins in both soluble and non-soluble fractions, showing that that the lysosomal function in treated autophagy-deficient GAA-/- mice is largely restored. It should be noted that even the most successful reversal of lysosomal pathology in ERT-treated autophagy-deficient GAA-/- mice leaves these animals autophagy-deficient in skeletal muscle. Our observational data (up to 18 months) have shown that skeletal muscle-specific suppression of autophagy in the wild type mice does not result in major abnormalities as shown by apparent strength, mobility, weight, and lifespan. Thus, the suppression of autophagy in skeletal muscle greatly facilitates the effect of ERT resulting in an outcome which has never been observed in Pompe mice with genetically intact autophagy. We have explored more deeply the regulation of autophagy in muscle cells. We have shown that glycogen synthase kinase-3beta - a protein involved in glycogen metabolism that has been shown to up-regulate autophagy in several other cell lineages is markedly dephosphorylated (activated) in Pompe mice, leading thereby to increased phosphorylation of glycogen synthase, which inactivates this enzyme. This may well be a regulatory response that has the effect of reducing the load of glycogen that is transported to lysosomes in Pompe disease, but it has the unfortunate consequence of up-regulating autophagy. We have also shown that GSK-3beta up-regulates autophagy in the muscle cell line C2C12 and in primary Pompe myoblasts. In collaboration Dr. Raben and Dr. Takikita, E Richard and G Douillard-Guillloux from France explored the consequences of substrate deprivation by using shRNA for glycogenin or for glycogen synthase in culture or delivering shRNA for glycogen synthase intramuscularly in mice using an AAV vector to Pompe mice. Both methods of blocking glycogen synthesis reduced glycogen accumulation in muscle cells. More recently, studies with Pompe mice lacking glycogen synthetase have been made and demonstrate that this method of substrate deprivation can prevent the clinical manifestations of Pompe disease. In yet another approach to rendering Pompe skeletal fast muscle fibers responsive to therapy, we have used the over-expression in muscle of the major regulatory protein, PGC1alpha, which changes some of the characteristics of fast muscle fibers into those of slow muscle fibers. We expected the PGC1alpha transgenic Pompe mice to have effective autophagy and respond well to ERT. The outcome was surprising. We found that the converted fast muscle cells did develop some characteristics of oxidative slow muscle fibers, and they lacked autophagic accumulation. However, these muscles developed a higher glycogen load and did not respond to ERT. Furthermore, they had markedly up-regulated numbers of lysosomes and autophagosomes, suggesting that PGC1alpha up-regulates the manufacture of these organelles as well as mitochondria, which had previously been shown by others. In collaboration with the NIAMS Laboratory of Muscle Biology, we have demonstrated that in Pompe fast muscle fibers, the maximum force is approximately half of that of comparable fibers from normal mice;that the normal hexagonal array of actin and myosin filaments is disordered;and that the fraction of actin-myosin cross bridges in the relaxed state is reduced. This points to the conclusion that the impaired myofibrillar function contributes to muscle weakness in Pompe disease. In a study of single muscle fibers from infants with Pompe disease - sent by colleagues in Taiwan - obtained before and after ERT, we have noted that the dominant pathology before therapy is the presence of huge, glycogen-filled lysosome but negligible autophagic build-up the hallmark of the milder adult form of the disease. However, with ERT, as the glycogen-filled lysosomes are shrunk, autophagic build-up becomes visible, as if unmasked by the clearance of glycogen. In most fibers the two pathologies did not seem to co-exist. The data point to the possibility of differences in the pathogenesis of Pompe disease in infants and adults, with possible implications for the design of therapy.